How your voice travels, how a channel is organized and what the app does to stay alert. Only what works today.
A note on honesty: audio is protected by TLS/WSS between your phone and our servers, but it is not end-to-end encrypted.
The microphone is sampled at 16 kHz and compressed with 4-bit IMA-ADPCM on the device itself, with no external libraries. Voice goes out in 100 ms frames, so what you say starts playing before you finish. Listeners keep a cushion of about 180 ms so the sound does not break up when the network wobbles.
Each channel hands out the floor: one person talks at a time and the service decides who holds it. Everyone else listens without answering, with an opening tone before each voice. You can talk privately to one person inside the channel, and the alert button notifies everyone or whoever you choose.
The connection between your phone and the service uses TLS and WSS. Your account is a MegaCity account: one identity for the whole ecosystem, and we never store your password here. Every payment and permission (who manages a channel, who sees the map) is validated on the server.
Voice comes from our own service, with no third-party voice platforms. For large operations, the Fleet plan includes a dedicated server in the city we agree on with you, to bring voice closer to your people and give your operation its own capacity. It is offered and set up on request.
The Android app keeps a foreground service while you have connected channels, receives private-voice and alert notifications with the screen off, and asks you to disable battery optimization so you do not miss alerts. On iPhone and computers you use the web version.
Sharing location is a decision of the person and of their team administrator. The app reads the GPS with whatever signal there is, slows down when you are still, and the server tells it whether to work in precise, normal or off mode. Routes carry stops, arrival time and the length of each visit; zones notify when someone enters or leaves.
| Voice codec | IMA-ADPCM, 4 bits per sample, 16 kHz |
|---|---|
| Data while talking | About 8 KB per second |
| Audio frames | 100 ms |
| Listening cushion | About 180 ms |
| Who talks | One person at a time per channel |
| Connection | TLS / WSS |
| Native app | Android (foreground service) |
| Other platforms | Installable web (PWA) |
| Languages | Spanish and English |
| Billing | Mexican pesos; prices in pesos or dollars |
Tell us about your operation and we will build the plan with you.
AI assistant: it can make mistakes. For anything important, write to us at /contacto.
